In collaboration with the laboratory of Jim Omichinsk,i we have identified a potential inhibitory domain of GATA-1 that reduces DNA binding to single GATA sites. Using peptides representing the zinc finger domain of GATA-1, we have mapped the activity to the N-terminal zinc finger of GATA-1. Tests with corresponding mutants of full length GATA-1 are in progress. In collaboration with Ritsuko Shimizu and Masi Yamamoto, we have shown that palindromic GATA sites are bound by both zinc fingers of a single GATA-1 molecule, while directly repeated sites are bound by the C-terminal fingers of two molecules of a GATA-1 dimer. 82% of solitary palindromic GATA-1 occupied sites in the genome show preferential binding by wild-type (WT) GATA-1 over a GATA-1 mutant in an amino acid necessary for N-finger DNA binding. WT GATA-1 induces a higher rate of transcription at pallindromic sites than does this mutant. EKLF is a transcription factor expressed somewhat later in erythroid development than GATA-1, and is also critical to this lineage. Anemias associated with mutations in the EKLF DNA binding domain have been identified in humans. An amino acid in the second zinc finger of EKLF is mutated in patients with congenital dyserythropoietic anemia (CDA) type IV and in the Nan (Neonatal anemia) mouse. Using purified EKLF zinc finger GST fusion proteins, in collaboration with Jim Bieker, we have confirmed that the anemic phenotype in the Nan mouse model is partially based on different DNA binding affinities of Nan-EKLF relative to wild type (WT), as suggested by earlier studies with nuclear extracts. In those studies it was shown that sites with a T base in the central triplet of the EKLF DNA binding motif were bound by WT EKLF, but not by Nan EKLF. Both WT and Nan EKLF bind sites with a C at this position. We have shown that the Kds for Nan binding to five of these T containing EKLF binding sites are higher than those for wild type EKLF. Thus one cause of the anemia in the Nan mouse is reduced binding affinity of Nan-EKLF for a subset of EKLF binding sites. Recently a new binding site for Nan-EKLF was discovered among genes that are up-regulated in Nan relative to WT mice. Many of these genes are usually expressed in macrophages, not erythroid cells. For two genes we have shown that the up-regulation is due to high affinity binding of Nan-EKLF to sites with this newly discovered motif to which WT EKLF does not bind. The protein products of these two target genes are secreted, and thus have cell- extrinsic effects that include inhibition of erythroid development. This finding partially explains the observation that while EKLF +/- heterozygous mice are normal, the Nan anemia occurs in heterozygotes in the presence of a normal copy of the EKLF gene. In collaboration with the laboratory of Jim Omichinski we have shown that p53 and the GATA-1 DNA binding domain interact through the transactivation domain of p53 and the linker and C-terminal zinc-finger of GATA-1. In collaboration with Masi Yamamoto, we plan to determine the role of this interaction during hematopoiesis by attempting to rescue GATA-1 null cells and mice with mutants of GATA-1 that do not interact with p53. Screens for GATA-1 mutants of this category are in progress.